
% Table created by stargazer v.5.2 by Marek Hlavac, Harvard University. E-mail: hlavac at fas.harvard.edu
% Date and time: Fri, Dec 06, 2019 - 11:27:26
\begin{table}[t]
  \begin{center}
    \caption{Comparison of Email Providers}
    \label{tab:email_compare}
    \begin{tabular}{@{\extracolsep{5pt}} rd{2}d{2}d{2}d{2}}
      \toprule
      \multirow{2}{1.25cm}{}
      & Registered & Provide & Assigned\textrm{ } A & Analytic \\
      & Voters       & Email    & Condition    & Sample   \\
      \midrule
      Proportion Democrat & $0.51$ & $0.52$ & $0.52$ & $0.51$ \\
      Age in 2016 & $51.52$ & $45.73$ & $45.50$ & $45.3$ \\ 
      Proportion Female & $0.53$ & $0.52$ & $0.52$ & $0.52$ \\
      Proportion non-White & $0.29$ & $0.31$ & $0.32$ & $0.30$ \\
      \hline \\ [-1.8ex]
      Observations & $12,339,702$ & $629,738$ & $503,859$ & $328,181$ \\
      \bottomrule
    \end{tabular}
  \end{center}
  {\linespread{1} \footnotesize \textit{Notes:} SEs are omitted here but reported in
    the SI.  \textit{Proportion Democrat} is the proportion of
    registered Democrats among those registered as a Democrat or
    Republican. %\textit{Likely Voter} is the proportion of individuals
    % who have voted in any election between 2012 and 2016.
    Column 2, \textit{Provide Email} includes registered voters with
    email addresses that could not be reconciled to be valid,
    \textit{Assigned A Condition}, includes only voters with 
    a valid email. The \textit{Analytic Sample}, used
    throughout the reported analysis, excludes voters who voted
    early in-person.} 
\end{table}
